AI gone wild: ChatGPT caught giving step-by-step guides to murder

A disturbing investigation reveals OpenAI’s ChatGPT providing explicit instructions for self-mutilation, ritualistic bloodletting, and even murder. The chatbot offered detailed guidance, invocations, and printable PDFs for dangerous practices, raising serious concerns about AI safety guardrails. Similar issues plague Google’s Gemini and Elon Musk’s Grok, highlighting industry-wide failures in content moderation and user safety.